The broad, long term objectives of this continuation grant are to define the natural history of the two major adult cholestatic liver diseases, primary biliary cirrhosis (PBC) and primary sclerosing cholangitis (PSC), by means of statistical models and to apply this knowledge to assess treatment efficacy and to optimize the timing of therapeutic intervention. At present, liver transplantation is the only therapy clinically accepted as lifesaving for these diseases. Because of improved transplantation survival rates during the past six years, randomizing patients into a nontransplantation (supportive treatment) control group has been considered clinically inappropriate. When developed, these statistical models can provide a mathematical control group to assess the efficacy of liver transplantation. In addition, statistical models will be developed to define risk factors and to predict outcome of liver transplantation. Another of the potentially important applications of the natural history model and the transplantation outcome model will be to improve the selection and timing of liver transplantation. Through such modeling, analyses will be performed to investigate whether patients in the "early-advanced stage" of liver disease should receive transplantation immediately or should delay receiving the procedure until the disease reaches the "late-advanced stage." The effect of transplantation on a patient's functional status, employability, and economic situation will be evaluated in addition to survival outcome. This proposal combines the strengths of two major medical centers, the Mayo Clinic and the University of Pittsburgh. At Mayo, NIH supported treatment trials have provided a unique, comprehensive, and longitudinal database for PBC and PSC patients without liver transplantation. This data base supports a novel approach to natural history modeling, which will describe the dynamic progression of the disease course as well as model survival outcome. The liver transplantation program at the University of Pittsburgh providws the largest single institutional experience in liver transplantation in the world. The Pittsburgh data, together with data from the Mayo transplantation program, provides outstanding resources for the development of the transplantation outcome models. These resources, together with the extensive clinical and statistical expertise in survival modeling at Mayo Clinic will provide answers to the critical issues of selection/timing and efficacy of liver transplantation.